Sip-Tools - Automated calls and answering machine

• sipcall - Automated calls over SIP/VOIP with TTS
• sipserv - Answering machine for SIP/VOIP with TTS

Dependencies:

• PJSUA API (http://www.pjsip.org)
• eSpeak (http://espeak.sourceforge.net)

Copyright (C) 2012 by Andre Wussow, desk@binerry.de

major changes 2017 by Fabian Huslik, github.com/fabianhu

more changes 2018 by Kaito Cross, github.com/KaitoCross

updated 2020 by Kat, github.com/KawaiiFiveO

For more information please visit http://binerry.de/post/29180946733/raspberry-pi-caller-and-answering-machine.

Installation on Raspberry Pi 2/3 with Raspian

1. Build and install PjSIP as explained below

2. install eSpeak sudo apt-get install espeak espeak-data

3. If output of DTMF digits through GPIO in sipserv is needed: (if not, skip to step 6)

4. install wiringPi as outlined on http://wiringpi.com/download-and-install/

5. after executing ./build as described in link, run sudo make install for wiringPi

6. Download this folder to Raspberry Pi

7. If output of DTMF digits in sipserv (answer machine) through GPIO is needed: execute make sipserv-gpio in SIP-Pi folder in Terminal

8. If forwarding the 4 last DTMF digits to a tcp server is needed, run make sipserv-tcp instead

9. add a -local directly at the end (without a space) if you run the sip server you are planning to connect to is on the same Pi. (Match any authentication realm request.)

10. make sipcall if you need the standalone phone software as well

11. If the mentioned output options are not needed, just execute make all in SIP-Pi folder in Terminal instead

12. Configure sipserv.cfg to your needs (see example configuration)

13. Test drive usingsudo ./sipserv --config-file sipserv.cfg

14. This is not(yet) a "real" service, so include ./sipserv-ctrl.sh start command into your favourite autostart.

15. Stop the SIP service using sipserv-ctrl.sh stop

16. Install lame sudo apt-get install lame for the MP3 compression of recordings (mail.sh)

sipserv

Pickup a call, have a welcome message played or read.
Do some actions by pressing (DTMF) keys on your phone.
Send 4-bit DTMF key value through GPIO if wished.
This service uses a generic approach. All actions are configurable via config file.
One special usage is the special ability to record the caller while playing the intro.
Please contact your lawyer, if this is legal in your country.
You can have a whitelist (numbers.txt) and only the special (whitelisted) calls answered.

Usage: (if GPIO output needed, has to run as root)

sipserv [options]

Commandline:

Mandatory options:

• --config-file=string Set config file

Optional options:

• -s=int Silent mode (hide info messages) (0/1)

Config file:

Mandatory options:

• ipv6=int IPv6 usage (0/1) Only set to 1 if you want to use IPv6 exclusively
• sd=string Set sip provider domain.
• su=string Set sip username.
• sp=string Set sip password.
• ln=string Language identifier for espeak TTS (e.g. en = English or de = German).
• call-log=string Filename to write call log to.
• tts=string String to be read as a intro message

determine DTMF Key digit encoding

• dtmf-encoding=int Set DTMF digit output binary encoding (0=linear/1=MT8870 scheme) (default linear)

and at least one dtmf key configuration (X = dtmf-key index):

• dtmf.X.active=int Set dtmf-setting active (0/1).
• dtmf.X.description=string Set description.
• dtmf.X.audio-response= Set audio response wav file to play; tts for that DTMF key will not be read, if this parameter is given. File format is Microsoft WAV (signed 16 bit) Mono, 22 kHz;
• dtmf.X.tts-intro=string Set tts intro.
• dtmf.X.tts-answer=string Set tts answer.
• dtmf.X.cmd=string Set shell command.

Optional options:

• rc=int Record call (0=no/1=yes)
• af=string announcement wav file to play; tts will not be read, if this parameter is given. File format is Microsoft WAV (signed 16 bit) Mono, 22 kHz;
• cmd=string command to check if the call should be taken; the wildcard # will be replaced with the calling phone number; should return a "1" as first char, if you want to take the call.
• am=string aftermath: command to be executed after call ends. Will be called with two parameters: $1 = Phone number $2 = recorded file name
• gpio-en=int enable output of DTMF digits on Raspberry Pi wiringPi GPIO

options for DTMF forwarding to TCP server:

• dtmf-value-forward-srv=string Set domain name of tcp server
• mail-audio-response=string _Set audio response for a processed email (in combination with dmail-connect only) (optional)

options for DTMF digit output on Raspberry Pi GPIO

The GPIO output function is based on wiringPi and uses the wiringPi numbering scheme.
It outputs the digits as 4-bit binary number.
When GPIO output has been enabled, you have to define all 4 output ports and the interrupt port.
To define them, put this into the config file:

• gpio-0=int Port number goes here instead of int
• gpio-1=int Port number goes here instead of int
• gpio-2=int Port number goes here instead of int
• gpio-3=int Port number goes here instead of int
• gpio-interrupt=int Port number goes here instead of int The interrupt port is often needed by the microprocesser that reads those digits.
  Connect it to the interrupt port of your other microprocesser and configure that controller as needed to respond to the interrupt.

a sample configuration can be found in sipserv-sample.cfg

sipserv can be controlled with

sudo ./sipserv-ctrl.sh start and
sudo ./sipserv-ctrl.sh stop

Changelog since fabianhu's version:

• Implemented IPv6 Support
• added option for wavefile being played instead of an tts file when DTMF key has been pressed
• added *, #, 0, and A-D into processable DTMF-signals
• added forwarding of last 4 DTMF digits to TCP server after # is pressed

Changelog since KaitoCross's version:

• sipserv updated and tested working on Raspbian Buster version September 2019
• Fixed various bugs and crashes
• Quality of life improvements, program status and error handling
• PjSIP installation instructions updated for compatibility

Build PjSIP

build directly on Raspberry Pi:

Note: SIP-Pi was only tested using PjSIP version 2.8.
Use newer or older versions at your own risk.

sudo apt-get install uuid uuid-dev libuuid1 libssl-dev libasound2-dev

cd ~/tmp # any installation directory
wget https://www.pjsip.org/release/2.8/pjproject-2.8.tar.bz2
tar xvfj pjproject-2.8.tar.bz2
cd pjproject-2.8/

edit pjlib/include/pj/config_site.h
Add next line into file:

#define PJ_HAS_IPV6 1

edit user.mak
Add following to file:

export CFLAGS += -march=armv7-a -mfpu=neon-vfpv4 -ffast-math -mfloat-abi=hard
export LDFLAGS +=

edit third_party/build/os-auto.mak.in
Edit near the end of the file and comment out the lines as follows:

# External webrtc
else
DIRS += webrtc
WEBRTC_OTHER_CFLAGS = -fexceptions -DWEBRTC_POSIX=1 @ac_webrtc_cflags@
#ifneq ($(findstring sse2,@ac_webrtc_instset@),)
#    WEBRTC_SRC = \
#    	      modules/audio_processing/aec/aec_core_sse2.o		 \
#	      modules/audio_processing/aec/aec_rdft_sse2.o	         \
#	      modules/audio_processing/aecm/aecm_core_c.o	         \
#	      modules/audio_processing/ns/nsx_core_c.o	                 \
#	      system_wrappers/source/cpu_features.o
#else ifneq ($(findstring neon,@ac_webrtc_instset@),)
WEBRTC_SRC = \
    [Modules, etc here]
WEBRTC_OTHER_CFLAGS += -DWEBRTC_HAS_NEON
#else ifneq ($(findstring mips,@ac_webrtc_instset@),)
#    WEBRTC_SRC = \
#              modules/audio_processing/aec/aec_core_mips.o               \
#	      modules/audio_processing/aec/aec_rdft_mips.o               \
#	      modules/audio_processing/aecm/aecm_core_mips.o             \
#	      modules/audio_processing/ns/nsx_core_mips.o                \
#	      common_audio/signal_processing/cross_correlation_mips.o    \
#	      common_audio/signal_processing/downsample_fast_mips.o      \
#	      common_audio/signal_processing/min_max_operations_mips.o
#
#    WEBRTC_OTHER_CFLAGS += -DMIPS_FPU_LE
#else # Generic fixed point
#    WEBRTC_SRC = \
#	      modules/audio_processing/aecm/aecm_core_c.o                \
#	      modules/audio_processing/ns/nsx_core_c.o                   \
#	      common_audio/signal_processing/complex_fft.o
#endif
endif
endif

(File ends here)

edit pjlib/include/pj/config.h
Replace this code:

    /*
     * ARM, bi-endian, so raise error if endianness is not configured
     */
#   if !PJ_IS_LITTLE_ENDIAN && !PJ_IS_BIG_ENDIAN
#       error Endianness must be declared for this processor
#   endif
#   if defined (PJ_M_ARMV7) || defined(ARMV7)
#	undef PJ_M_ARMV7
#	define PJ_M_ARM7		1
#	define PJ_M_NAME		"armv7"

with this:

#   define PJ_IS_LITTLE_ENDIAN  1
#   define PJ_IS_BIG_ENDIAN     0
#   if defined (PJ_M_ARMV7) || defined(ARMV7)
#	undef PJ_M_ARMV7
#	define PJ_M_ARM7		1
#	define PJ_M_NAME		"armv7"

navigate back to pjproject-2.8 folder

./configure --disable-video --disable-libwebrtc
make dep
make
sudo make install

You will have plenty of time to drink some Dr. Pepper during make. Enjoy while waiting.

Cross build of PjSIP for Raspberry:

Legacy instructions. Not tested on new versions of Raspbian.

export CC=/opt/raspi_tools/tools/arm-bcm2708/gcc-linaro-arm-linux-gnueabihf-raspbian-x64/bin/arm-linux-gnueabihf-gcc
export LD=/opt/raspi_tools/tools/arm-bcm2708/gcc-linaro-arm-linux-gnueabihf-raspbian-x64/bin/arm-linux-gnueabihf-gcc
export CROSS_COMPILE=/opt/raspi_tools/tools/arm-bcm2708/gcc-linaro-arm-linux-gnueabihf-raspbian-x64/bin/arm-linux-gnueabihf-
#export AR+=" -rcs"

export LDFLAGS="-L/opt/raspi_tools/tools/arm-bcm2708/gcc-linaro-arm-linux-gnueabihf-raspbian-x64/lib/gcc/arm-linux-gnueabihf/4.8.3 -L/opt/raspi_tools/tools/arm-bcm2708/gcc-linaro-arm-linux-gnueabihf-raspbian-x64/arm-linux-gnueabihf/lib -ldl -lc"

./aconfigure --host=arm-elf-linux --prefix=$(pwd)/tmp_build --disable-video --disable-libwebrtc

Add into pjlib/include/pj/config_site.h:
#define PJ_HAS_IPV6 1

navigate back to pjproject-2.8 folder

make dep

make

sipcall

Make outgoing calls with your Pi. Not tested on new versions of Raspbian.

Usage:

• sipcall [options]

Mandatory options:

• -sd=string Set sip provider domain.
• -su=string Set sip username.
• -sp=string Set sip password.
• -pn=string Set target phone number to call
• -tts=string Text to speak

Optional options:

• -ttsf=string TTS speech file name
• -rcf=string Record call file name
• -mr=int Repeat message x-times
• -s=int Silent mode (hide info messages) (0/1)
• -ipv6=int IPv6 mode (use IPv6 exclusively) (0/1)

see also source of sipcall-sample.sh

License

This tools are free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation; either version 2.1 of the License, or (at your option) any later version.

This tools are distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details.